#emc | Logs for 2005-11-30

[00:09:33] <jepler> skunkworks: ^^^
[00:12:08] <skunkworks_emc2> Well now I have a gray cone - sure doesn't look as pretty ;) thanks
[00:12:18] <skunkworks_emc2> do you think it is my system or a but in axis?
[00:12:27] <skunkworks_emc2> bug
[00:14:35] <jepler> before you made that change, the top of the cone was the same color as the side -- both were pure black?
[00:15:43] <jepler> do you have the program "glxgears" installed on your system? It's another OpenGL program that displays some lit surfaces. See whether it looks right.
[00:17:57] <Jymmm> Jymmm is now known as Red70sShow
[00:17:57] <Red70sShow> Red70sShow is now known as Jymmm
[00:18:03] <skunkworks_emc2> you could not see the cone - just what it passed oever
[00:18:05] <skunkworks_emc2> over
[00:18:55] <skunkworks_emc2> where would I find the glxgears?
[00:19:53] <jepler> on this fedora machine and the bdi machine at work, it was installed by default
[00:20:01] <jepler> fedora reports that it's part of "xorg-x11-tools"
[00:20:25] <jepler> I can't access my bdi machine right nowu
[00:20:45] <skunkworks_emc2> your starting to talk greek again. is it something I can run from the command line or is it a menu item
[00:20:58] <jepler> "glxgears" is something you would type at the commandline
[00:21:51] <skunkworks_emc2> localhost:~# glxgears
[00:21:51] <skunkworks_emc2> Error: couldn't open display (null)
[00:21:51] <skunkworks_emc2> localhost:~#
[00:22:03] <skunkworks_emc2> wait
[00:22:41] <skunkworks_emc2> yes that works
[00:23:01] <skunkworks_emc2> can't run it from root ;) - I am learining
[00:23:09] <jepler> so the different parts of the gears are lighter or darker, like a light was shining from one side
[00:23:20] <skunkworks_emc2> yes
[00:23:32] <skunkworks_emc2> parts get lighter as they spin
[00:23:41] <skunkworks_emc2> pretty cool
[00:24:01] <skunkworks_emc2> around 15 frames per second
[00:25:47] <jepler> I was hoping it would be black too, because then my code would be found faultless
[00:26:10] <jepler> If I ask you to try applying a patch, do you know what I mean or am I talking geek again?
[00:27:23] <skunkworks_emc2> If you can walk me though it I can do it.
[00:28:46] <jepler> first, use a web browser to download this file to the directory /tmp: http://emergent.unpy.net/index.cgi-files/sandbox/axis-lights.patch
[00:29:01] <jepler> then, at the commandline, go to the axis source directory
[00:29:12] <jepler> then type this command: patch -p0 < /tmp/axis-lights.patch
[00:29:24] <jepler> you also have to undo the changes you made earlier
[00:29:41] <jepler> let me know if you get a lit cone or still get a black one
[00:30:21] <skunkworks_emc2> I just did undo the remarking out of the glEnable(GL_LIGHTING) command and it is black again
[00:30:47] <jepler> did you do the "patch" yet?
[00:30:57] <skunkworks_emc2> hold on - hold on ;)
[00:31:01] <jepler> OK OK
[00:34:36] <skunkworks_emc2> Ok I patched now do I - sudo env EMCROOT=/home/sam/emc2 python setup.py install
[00:35:24] <jepler> yes, that's the next step, unless patch told you that it FAILED in some way
[00:35:30] <skunkworks_emc2> sam@localhost:~/axis-ps249$ patch -p0 < /tmp/axis-lights.patch
[00:35:30] <skunkworks_emc2> patching file extensions/minigl.c
[00:35:30] <skunkworks_emc2> patching file rs274/OpenGLTk.py
[00:36:26] <jepler> if this doesn't yep, go on with "setup.py install"
[00:36:30] <jepler> er
[00:36:35] <jepler> yep, go on with "setup.py install"
[00:40:54] <skunkworks_emc2> that seemed to work - did you fix that just now? what was the problem? Your good
[00:42:39] <jepler> will you be disappointed if I tell you I'm not sure what the problem is?
[00:43:41] <skunkworks_emc2> I believe you - I have those kinda problems too.
[00:45:07] <jepler> My guess was basically that AXIS was assuming that OpenGL would behave in a certain way by default. axis-lights.patch makes sure that some appropriate settings are made.
[00:47:08] <jepler> but for me the change also made the cone a little brighter overall
[00:56:11] <skunkworks> Got disconnected - the cone looks good to me - pretty dark though. but i like it
[00:56:22] <skunkworks> is there a way to change the jog feed rate?
[01:03:30] <skunkworks> am I getting out?
[01:04:31] <Jymmm> skunkworks: No, speak close to the mic
[01:05:01] <skunkworks> ;) switched computers
[01:05:08] <Jymmm> =)
[01:05:24] <skunkworks> is there a way to change the jog feed in axis?
[01:05:42] <Jymmm> * Jymmm doesn't know, but would expect so.
[01:06:38] <skunkworks> Ok - I just don't see it I am sure.
[01:16:07] <skunkworks> jepler - thanks again
[01:17:13] <jepler> skunkworks: the feed override slider also controls the jog speed
[01:17:21] <skunkworks> ah
[01:17:34] <jepler> skunkworks: and you're welcome. When I can fix a bug, I'm thrilled.
[01:17:35] <skunkworks> I will give it a try
[01:18:03] <skunkworks> is that patch going to be added to the download any time soon?
[01:19:05] <jepler> skunkworks: the cone fix? Yes, later tonight.
[01:19:31] <skunkworks> cool
[01:20:07] <jepler> in fact it should already be in http://axis.unpythonic.net/index.cgi-files/downloads/nightly/axis-latest.tar.bz2
[01:21:19] <jepler> see the description of patchset 250 on this page. http://axis.unpythonic.net/index.cgi/downloads/nightly
[01:21:33] <skunkworks> I just downloaded the file about 2 hours ago
[01:22:47] <jepler> patchset 250 was added at 19:44:36 UTC, or about half an hour ago.
[01:22:56] <jepler> er, I guess that must be local time
[01:23:01] <jepler> I thought those logs showed UTC
[01:24:24] <skunkworks> ah - I think I am little slow tonight - My name is on it. I feel special.
[01:27:40] <jepler> hm. that "nightly" page looks like crap on a 1024-wide monitor, because of how long some of the lines are in the changelog...
[01:27:47] <jepler> I should do something about that
[01:30:24] <jepler> ah, that's an improvement
[01:30:35] <jepler> too bad about the HTML
[01:38:21] <jepler> well it looks prettier anyhow
[01:45:30] <skunkworks> Do you think it is an intesity problem with my computer? the cone is shaded now but pretty dark - when the light source was behind - it was totally dark.
[01:48:58] <jepler> It's possible. The darkest part of the cone should now be about 7.5% as bright as the brightest pixel your monitor can display, if I understand the way opengl calculates lighting.
[01:51:21] <jepler> in the file rs274py/OpenGLTk.py, below 'def basic_lighting()', you might try changing the numbers on the GL_AMBIENT and GL_DIFFUSE lines. Increasing GL_AMBIENT will increase the brightness of the darkest parts. The numbers should sum to 1. You could try .4 and .6, say. After you change something, "setup.py install" and restart emc to see if you like the results.
[01:51:54] <jepler> (the 4th number should always be 1)
[01:52:14] <skunkworks> sounds good - now that I am finding my way around. I will try that tomarrow.
[01:52:20] <jepler> OK
[02:06:43] <skunkworks> Good night
[02:07:59] <jepler> good night
[02:35:08] <CIA-5> 03jmkasunich * 10emc2/src/hal/drivers/hal_ppmc.c:
[02:35:09] <CIA-5> Changed calculation of stepgen divisor to reflect the fact that the hardware
[02:35:09] <CIA-5> actually divides by the specified number plus 4. The error at normal step rates
[02:35:09] <CIA-5> is small and inside a closed loop, so it doesn't effect the machine, but it was
[02:35:09] <CIA-5> still wrong.
[02:41:22] <jepler> "affect"
[02:41:33] <jmkasunich> fsck it
[02:41:45] <jmkasunich> I actually thought about that, I always get those two mixed up
[02:41:49] <Jymmm> * Jymmm hands jmkasunich a beer or 20
[02:42:41] <jepler> hah
[02:42:47] <jepler> I enjoy being a grammar prick now and then, that's all
[02:43:47] <Jymmm> * Jymmm looks at jepler and shakes his head
[02:44:36] <SWPadnos_> the fact that it has no effect doesn't affect the machine
[02:46:22] <jmkasunich> your grammer affectations will have no effect on me!
[02:47:20] <Jymmm> * Jymmm arssssssssssss kollage edjoomakated
[02:48:05] <jmkasunich> 5 years ago I cudn't spel inguneer, and now I ar one!
[02:48:06] <SWPadnos_> ain't I never told you not to get no bad grades in grammer?
[02:48:14] <jepler> I quit college
[02:48:59] <SWPadnos_> me too
[02:49:45] <jepler> it looks like there's some exciting stuff in firefox 1.5.
[02:51:21] <jepler> maybe we'll call the next AXIS 1.5 just to keep up with mozilla's version numbers
[02:51:48] <jmkasunich> does that mean you can't do a followup release until they do?
[02:52:50] <jepler> oh they have bug fixes all the time
[02:53:45] <jmkasunich> so do you ;-)
[02:53:50] <jmkasunich> but actual releases?
[02:54:41] <jepler> well we don't have bugs in axis
[02:54:47] <jepler> we just have features I didn't quite finish implementing
[02:54:54] <jmkasunich> you have featiures
[02:54:58] <jmkasunich> features
[02:55:08] <jmkasunich> (some of which are negative features)
[02:55:28] <jepler> that's another way to look at ti
[02:55:31] <jepler> goodnight folks
[02:55:38] <jmkasunich> night
[03:01:10] <Jymmm> G'night
[03:01:33] <Jymmm> we need to make a CNC needle for jepler's ego =)
[03:49:25] <Jymmm> hard hard is it to turn 1/8" to 1/4" diameter brass, SS ?
[03:56:47] <cradek> it's impossible without a "putting-on" tool
[03:57:09] <Jymmm> what is that?
[03:57:25] <cradek> it's what you use when you want to enlarge a diameter on the lathe
[03:57:52] <Jymmm> * Jymmm wonders if he needs to put on his boots
[03:58:16] <cradek> kidding aside, what did you mean?
[03:58:38] <Jymmm> cradek: Like making decrotive cribbage board pegs
[03:59:01] <cradek> with small diameters you get deflection
[03:59:09] <cradek> other than that, there's no special problem
[03:59:33] <cradek> if you don't stick the material very far out of the chuck, and you use a tailstock, it's probably fine
[04:00:03] <cradek> brass cuts easily but doesn't stay very nice looking - it tarnishes
[04:00:27] <cradek> stainless can be hard to cut with a nice finish
[04:00:53] <Jymmm> ok, china bulk sounds better right about now =)
[04:01:09] <cradek> aluminum is a nice silver metal that cuts easily and doesn't tarnish
[04:01:15] <cradek> it might feel cheap though, being pretty light
[04:01:42] <cradek> I'm not sure what shape cribbage board pegs have
[04:02:20] <Jymmm> nothing fancy, usually just different colors per player
[04:02:42] <Jymmm> a shaft with chunk 'handle'.
[04:03:47] <cradek> how about polished cylinders with a nice knurl on top? no cutting required
[04:04:00] <cradek> can't knurl stainless
[04:04:05] <cradek> brass and aluminum are easy
[04:04:14] <Jymmm> how do you knurl?
[04:04:18] <cradek> well, *I* can't knurl stainless
[04:04:30] <cradek> push the knurling tool against the slowly rotating work
[04:04:33] <Jymmm> you mean criss cross pattern?
[04:04:37] <cradek> right
[04:04:40] <Jymmm> ah, ok.
[04:04:41] <cradek> hatching for gripping
[04:04:53] <Jymmm> yeah, I get it now.
[04:04:56] <cradek> some knurling tools "pinch" the work
[04:05:06] <cradek> (tighten from both sides)
[04:05:44] <Jymmm> is this somethign that requires a lathe?
[04:06:07] <cradek> yeah, probably
[04:06:16] <cradek> maybe you could rig something but a lathe is the obvious tool to use
[04:06:20] <cradek> http://www.sherline.com/3004inst.htm
[04:06:29] <cradek> this is sherline's knurling tool that pinches the work
[04:07:29] <Jymmm> Ok, this is getting complicated =)
[04:07:36] <cradek> just a thought
[04:07:39] <cradek> you don't have a lathe?
[04:07:39] <Jymmm> * Jymmm slowly backs up =)
[04:07:48] <Jymmm> no, just my router
[04:07:55] <cradek> it's very hard to make cylindrical parts without a lathe
[04:08:00] <cradek> very easy to make them with one!
[04:08:09] <Jymmm> I got a drill press =)
[04:08:42] <cradek> you could probably knurl in a drill press IF it's a good one that can be configured to run VERY slow
[04:09:00] <Jymmm> Hmmm. let me check....
[04:09:01] <cradek> like you would use with very large diameter drills
[04:09:26] <Jymmm> 540-3600 RPM
[04:09:49] <cradek> I'm trying to picture 540
[04:09:56] <cradek> that's probably way way too fast to knurl
[04:09:56] <Jymmm> it's fast
[04:10:14] <cradek> cheap drill presses don't do slow speeds well
[04:10:23] <cradek> really limits your max drill size
[04:10:51] <Jymmm> It's served me well so far =)
[04:11:15] <cradek> I bet you've never drilled over a 1/2" hole with it
[04:11:24] <Jymmm> 6"
[04:11:35] <Jymmm> w/ fly cutter
[04:11:42] <cradek> scaaaary
[04:11:47] <cradek> crap, I should go to bed
[04:12:07] <Jymmm> cradek anyone to keep the bed warm?
[04:12:18] <cradek> noyfb
[04:12:50] <Jymmm> By that comment I'd say.... Get the fuck out of here and go to bed!
[04:13:02] <cradek> ha
[04:13:17] <cradek> I usually sleep with three others, two of which are cats
[04:13:59] <Jymmm> * Jymmm dives off the cliff wiht the following comment.....
[04:14:13] <Jymmm> Man that's a lot of pussy to be sleeping with
[04:14:31] <cradek> ha
[04:14:52] <cradek> you assume a lot
[04:15:10] <cradek> anyway, goodnight
[04:15:10] <Jymmm> well, assuming you're not gay that is.
[04:15:13] <Jymmm> G'Night
[04:39:17] <CIA-5> 03paul_c * 10emc2-auto/wiki/ (17 files in 12 dirs): "Auto update wiki from a cron job. Wed Nov 30 05:30:01 GMT 2005 "
[06:13:58] <SWPadnos_> SWPadnos_ is now known as SWP_Away
[07:48:31] <CIA-5> 03jmkasunich * 10emc2/ (configs/ppmc_motion.hal src/hal/drivers/hal_ppmc.c): Added an enable pin to the PPMC step generators, and updated the .hal file to use it. Added driver code for the PWM outputs on the Universal PWM Controller, but can't test yet (no hardware).
[08:37:39] <K`zan> Night folks
[11:29:05] <Jacky^> morning
[12:13:27] <cradek> hello
[12:13:38] <Jacky^> hi cradek
[12:14:09] <alex_joni> morning guys
[12:14:18] <cradek> alex_joni: feeling any better?
[12:14:21] <Jacky^> mornin alex_joni
[12:14:32] <alex_joni> cradek: a bit, but I stayed home
[12:14:36] <cradek> alex_joni: everyone around here is sick too
[12:14:57] <cradek> good, that'll fix you up
[12:15:44] <alex_joni> hope so ;)
[12:15:47] <alex_joni> got lots to do :)
[13:23:37] <lerman> I've got some sort of cold, also. I think I caught it from you guys. :-)
[13:23:51] <les_w> morning all
[13:24:53] <cradek> hi all
[13:25:17] <alex_joni> lerman: sorry to hear that
[13:25:23] <alex_joni> morning les
[13:25:43] <les_w> partly an office day for me
[13:26:06] <alex_joni> it's been a day in bed for me ;)
[13:26:17] <les_w> oh I saw you were sick
[13:26:19] <les_w> sorry
[13:26:28] <alex_joni> getting better, slowly
[13:26:39] <les_w> I'm ok and the weather is nice
[13:27:03] <les_w> might have to take off and hit the golf course
[13:27:19] <les_w> if it gets warm enough
[13:28:11] <les_w> right now I am just getting an order together for MSC
[13:28:27] <les_w> some of that nice igus hiflex cable
[13:29:26] <les_w> oops just remembered...it's cheaper direct from igus
[13:29:30] <les_w> I think
[13:39:37] <les_w> I can't understand why their vfd/servo cable has 4 conductors minimum (plus shield)
[13:39:51] <les_w> for open delta I only need 3
[13:40:29] <les_w> I guess wye would need 4
[13:48:22] <crib> re
[14:11:33] <alex_joni> morning ray
[14:11:58] <rayh> Hi Alex
[14:14:16] <jepler> It looks like AXIS needs to determine what extra M-codes (M101 etc) are available so that it can provide a preview of files that use them. Anyone know if there's an existing way to get this information out of emc?
[14:14:51] <alex_joni> jepler: yes
[14:15:03] <alex_joni> there is a piece of software that looks for the M-codes
[14:15:08] <alex_joni> and adds those to a list
[14:15:13] <rayh> phone brb
[14:15:22] <alex_joni> some m-code list something, can't really recall the name
[14:15:28] <alex_joni> but it exists in the interpreter
[14:15:50] <jepler> Yeah, I know that it's shared between some part of emc and the interpreter, but is it available to the GUI?
[14:15:54] <alex_joni> not sure how you get that to the GUI..
[14:16:04] <alex_joni> why do you need it in the GUI again?
[14:16:12] <alex_joni> s/again//
[14:17:49] <jepler> otherwise AXIS can't give the preview plot. Given this program on 3 lines, AXIS will just show an error instead of showing a linear move: M101 / G1 X1 / M2
[14:18:41] <SWP_Away> SWP_Away is now known as SWPadnos_
[14:18:55] <alex_joni> I see
[14:19:04] <alex_joni> how about ignoring Mxx ?
[14:19:18] <SWPadnos_> unfortunately, the external M-codes are actual programs - there's no way to do a dry run with them
[14:19:20] <alex_joni> do they make a difference on the preview?
[14:19:38] <alex_joni> SWPadnos_: not dry run with them..
[14:19:58] <SWPadnos_> a "see what this would do" run :)
[14:20:01] <jepler> In the preview, I'd like M101 to do nothing, or maybe show a dwell-like symbol at that location.
[14:20:02] <alex_joni> jepler: is Mx the way you figure out between moving and machining?
[14:20:19] <alex_joni> I'd ignore M1xx for now
[14:20:37] <jepler> but if M103 isn't defined, I want to correctly tell the user that the program is sure to abort when run
[14:20:46] <alex_joni> I see
[14:20:54] <jepler> so I need to know the list of available external M-codes
[14:22:07] <rayh> jepler: Have you tested with Lerman's code?
[14:22:32] <jepler> rayh: no, I haven't.
[14:22:34] <alex_joni> jepler: active_m_codes
[14:22:56] <jepler> rayh: some things he told me makes me think there will be problems to iron out
[14:23:10] <alex_joni> jepler: the GOOD news, it should already be in EMC_TASK_STAT
[14:23:17] <jepler> alex_joni: no, active_m_codes is something else
[14:23:26] <alex_joni> ahh.. the ones active right now
[14:23:50] <SWPadnos_> M100+ aren't modal
[14:24:11] <SWPadnos_> nevermind - you knew that
[14:24:42] <ValarQ> hello folks
[14:24:51] <ValarQ> hello rayh
[14:24:55] <lerman> What DO the M100+ codes do? (generally) What are they good for?
[14:25:41] <ValarQ> rayh: you was asking something about the halgui?
[14:25:55] <jepler> lerman: they can do anything you like. I'm not sure what people use them for.
[14:26:04] <rayh> phone again brb
[14:26:09] <lerman> If they are just used for generic machine control, I would think they could be replaced by a combination of M codes that interface with HAL, together with some g code.
[14:26:18] <SWPadnos_> I don't think people use them much (obviously not with emc2, because they don't work :) )
[14:26:36] <jepler> well for now I think I'll go the easy route and accept all M100-M199
[14:26:41] <alex_joni> = { 0 };
[14:26:57] <alex_joni> USER_DEFINED_FUNCTION[] is the array that holds the data
[14:27:09] <lerman> I wouldn't think you would use them to (e.g., machinine a curve or pocket). They'd be more likely to be used for turning on a special coolant line, etc.
[14:27:12] <alex_joni> it shouldn't be too hard to get that to NML somehow..
[14:28:54] <rayh> Tomp tested a m104 to turn on a hal pin
[14:29:03] <rayh> Worked great
[14:29:09] <alex_joni> nice
[14:29:16] <rayh> halcmd
[14:29:27] <alex_joni> sets ?
[14:29:47] <rayh> Yep. Not as good a solution as nml but quick
[14:29:52] <lerman> As far as I can tell, ALL interfaces to hardware should be through HAL. In that case, codes to read or write to HAL should be all that is needed. HAL can write a value or read a value, write a bit, or read a bit. If those things are already accessible, then you don't need the ability to run a script or program. (Although that doesn't mean that some people might not prefer that capability.)
[14:30:23] <alex_joni> lerman: you still need the hooks to call that code
[14:30:33] <alex_joni> to map Mxx to a certain HAL action
[14:30:41] <rayh> The only thing I'd like to add to TomP's work would be a "wait done" message.
[14:30:41] <alex_joni> and that is easier with a script
[14:30:49] <jepler> lerman: how ironic if you think there's something in the interpreter that is a bad idea
[14:31:09] <lerman> Correct. YOu would need one MXX to set a pin, one to read a pin, one to write a value, one to read a value.
[14:31:22] <alex_joni> sounds good to me
[14:31:44] <alex_joni> and have them userdefineable, not M123 = set pin, M124 = read pin
[14:31:45] <lerman> No, not a bad idea. Just one that might be superceded by other capabilities.
[14:32:10] <rayh> Tom wrote a bash script and named it m104
[14:32:27] <SWPadnos_> it would be silly to use an external script or program to just set a pin
[14:32:36] <cradek> Mxxx is not good enough for coordinated IO
[14:32:37] <SWPadnos_> that was a proof of concept, not a suggestion of usage
[14:32:42] <lerman> Why would they need to be user definable? The user would write a subroutine that used a generic M123 to set a pin.
[14:33:09] <rayh> Morning Bob.
[14:33:16] <websys> Hello Ray
[14:33:35] <websys> How much snow are you under up there?
[14:33:36] <SWPadnos_> the M codes get two parameters as well - you could specify that you want to set the 17th signal to 43 if you want, but that's silly
[14:33:52] <lerman> Well, you coordinate IO by having an MXX (or a switch on one) that caused a wait until all pending IO had been completed. Think of it as a 'flush' buffer command.
[14:33:52] <rayh> Snow was washed away by the rain.
[14:34:04] <jepler> I have never used M1xx myself, but it seems like there are any number of uses, not limited to setting HAL signals
[14:34:05] <alex_joni> websys: hi bob
[14:34:15] <websys> Howdy alex
[14:34:25] <alex_joni> jepler: right, you might have M102 -> rm -rf /*
[14:34:33] <alex_joni> you are running as sudo ;)
[14:34:46] <cradek> lerman: imagine you don't want the machine to stop, you just want to turn something [the spray paint nozzle] on, coordinated with motion
[14:34:47] <rayh> I like that one alex.
[14:35:01] <alex_joni> rayh: yeah I know, you pointed it out sometime too
[14:35:12] <cradek> I'd like to say again that running emc as sudo is stupid.
[14:35:21] <SWPadnos_> or m106 -> wget "newparameterfile" / load_parameters
[14:35:21] <alex_joni> cradek: coordinated with motion is another thing
[14:35:35] <rayh> I even used it once. Had a box that needed rebuilding. That is a tested ability.
[14:35:42] <alex_joni> cradek: but it might not need be
[14:35:50] <lerman> To coordinate: Mxxx wait for IO to complete, Myyy turn on coolant.
[14:35:54] <cradek> alex_joni: we had m64,m65 in emc1
[14:35:59] <websys> needed rebuiling after the rm -rf /* ?
[14:36:02] <alex_joni> cradek: yeah, those will be back in emc2
[14:36:03] <cradek> alex_joni: those need to be halified somehow I'm sure
[14:36:25] <alex_joni> websys: not after, before ;)
[14:36:33] <alex_joni> websys: that was the first step
[14:36:37] <rayh> Na it needed it before the rm -rf but there was nothing left.
[14:36:41] <rayh> after.
[14:36:57] <websys> very efficient command
[14:37:16] <alex_joni> I imagine a fdisk command would do that quicker ;)
[14:37:17] <rayh> Interesting to watch the system sink into the mud.
[14:37:22] <alex_joni> back to M64,65
[14:37:41] <alex_joni> cradek: I talked to jmk a while back, and the code is only commented out in emc2
[14:37:45] <SWPadnos_> dd if=/dev/random of=/dev/hda
[14:37:46] <rayh> Are m64 .. tested to coordinate.
[14:37:50] <alex_joni> not VERY much work to do that
[14:38:06] <cradek> alex_joni: the implementation is specific to parport I think
[14:38:19] <alex_joni> in emc1
[14:38:27] <lerman> gotta go. see you later. Ken
[14:38:34] <cradek> rayh: yes I understand they coordinate to motion
[14:38:43] <alex_joni> it will be a number of 2 pins exported from the motion controller, and the integrator can connect them to anything
[14:38:43] <SWPadnos_> there should be HAL pins for the aux outs
[14:38:59] <alex_joni> one pin for DIO, one for AIO
[14:39:02] <rayh> Exactly. A 64 pin.
[14:39:11] <SWPadnos_> an M64 pin ;)
[14:39:18] <alex_joni> a 64 pin for DIO, and 65 for AIO
[14:40:52] <rayh> Good plan. That makes that implementation easy.
[14:41:13] <alex_joni> oh my.. did you guys look at http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?EmcFeatures ?
[14:41:55] <SWPadnos_> uh-oh - who let qconfig out ;)
[14:42:27] <rayh> Uh hey. That's what a wish list is all about.
[14:42:44] <rayh> "please santa don't be late!"
[14:42:50] <alex_joni> see swampy, we didn't do that for nothing :)
[14:43:07] <SWPadnos_> yeah - I'm now world famous in the emc wiki :)
[14:43:32] <SWPadnos_> of course, Ray gets all the credit
[14:43:53] <rayh> * rayh hides
[14:44:09] <alex_joni> SWPadnos_: that's how it usually works ;)
[14:44:20] <SWPadnos_> I don't mean for real - I mean that the wiki says yhis "the basic setup, namely for "hobby machines", is still too complicated. A solution like Ray's (?) unfinished "emc-qconfig" would be helpfull:"
[14:44:24] <SWPadnos_> ;)
[14:45:18] <alex_joni> yeah I know ;) .. that's why I said it usually works like that
[14:45:26] <SWPadnos_> heh - politics
[14:45:29] <alex_joni> you do all the hard work, then some suit gets the credit :D
[14:45:34] <rayh> I spent a few hours on a tickle parser for halcmd show last night.
[14:45:35] <alex_joni> rayh: no offence
[14:45:40] <SWPadnos_> ray == suit - hmmm
[14:45:48] <alex_joni> no.. rayh != suit :)
[14:45:51] <rayh> That is damn awkward
[14:45:57] <jepler> ray - suit == argh my eyes
[14:45:59] <alex_joni> rayh: what is?
[14:46:06] <jepler> put some clothes on!
[14:46:29] <rayh> I had one of those suit things, and a penguin tie but that was long ago
[14:46:40] <alex_joni> penguin? as in .. tux ?
[14:46:41] <alex_joni> :D
[14:46:46] <SWPadnos_> sorry - I talked to jmk about making the halcmd changes we had discussed, and he asked me to wait for some change the was thinking about
[14:46:51] <rayh> Tux rules.
[14:47:05] <SWPadnos_> /the/he/
[14:47:18] <rayh> I wore it into a used computer store and the manager was pissed.
[14:47:29] <rayh> pardon my expletive
[14:47:42] <SWPadnos_> would P-O'ed be PC?
[14:48:10] <rayh> IMO we need to get lerman's changes into main.
[14:48:19] <rayh> along with m1xx
[14:48:37] <rayh> and the m64-5 stuff as well
[14:49:09] <rayh> I fear for the consequences to that for the gui's.
[14:49:17] <SWPadnos_> incidentally, Gene Heskett mentioned the G-code reference in Machinery's Handbook - that's probably a good place to look for extensions to Kramer's doc
[14:49:56] <rayh> lerman said that he had a cam subroutine working.
[14:50:19] <rayh> Seems to me that if we can handle these sorts of expansions as subroutines
[14:50:30] <rayh> we don't have to do them in the interpreter itself.
[14:50:38] <SWPadnos_> well - lerman is a programmer by nature - he has an advantage over the average machinist
[14:51:03] <alex_joni> right, but the average machinist might take your bad-ass qconfig thingie
[14:51:09] <rayh> Yes. But the ellipse code from MH isn' that complex
[14:51:11] <alex_joni> and select the subroutine he needs :D
[14:51:27] <SWPadnos_> yeah - bad-ass machining RAD, man!
[14:51:38] <rayh> If we have a repository for subroutines, and have them documented.
[14:51:52] <SWPadnos_> yeah - that would be a good thing
[14:52:03] <rayh> linuxcnc.org can handle that.
[14:52:17] <SWPadnos_> maybe there should be a "subs" directory under nc_files
[14:52:28] <SWPadnos_> put in a couple of examples
[14:52:32] <rayh> We could access the actual file from the wiki documentation.
[14:52:44] <les_w> hi ray. a question....?
[14:52:53] <rayh> I'd rather not see them go with the code itself.
[14:53:12] <rayh> That way we don't have to take guff when one doesn't work.
[14:54:19] <SWPadnos_> I wouldn't include all of them - but like M101 and M102, include a couple to show how it's done
[14:55:14] <rayh> Yes. For these M1xx things we should put up a wiki page.
[14:55:31] <rayh> In fact I'd like to see a master wiki page for interp.
[14:56:20] <SWPadnos_> there's a GCode page, but you have to search for it (actually two of them)
[14:57:24] <les_w> Time to wire up the spindle vfd. Speed can be programmed to respond to a 0-10v analog signal. Does anyone have .ini examples of this?
[14:57:41] <rayh> les_w: You had a question
[14:58:24] <rayh> On your stg card?
[14:58:31] <les_w> yeah
[14:58:44] <rayh> looking
[14:58:49] <les_w> thanks
[14:59:37] <les_w> it's a little scary to have the pc control a big spindle....must use plenty of safety switches
[15:00:44] <rayh> I suspect that paul was using the parport but there is an example in the lathe stuff BDI-30
[15:00:50] <rayh> # analog out bits
[15:00:50] <rayh> SPINDLE_ON_INDEX = 3
[15:00:50] <rayh> MIN_VOLTS_PER_RPM = -0.001
[15:00:51] <rayh> MAX_VOLTS_PER_RPM = 0.001
[15:01:53] <les_w> hmm
[15:02:08] <les_w> why a min and max volts per rpm?
[15:03:29] <les_w> I just need 10v=24000 rpm, 0v= 0 rpm
[15:03:32] <rayh> Wah! Look at vital.ini in EMC.
[15:03:46] <SWPadnos_> if it can only go forward, set MIN to 0 ??
[15:04:22] <alex_joni> SWPadnos_: then it would be POS_VOLTS and NEG_VOLTS
[15:04:22] <les_w> can go reverse, but don't plan to much
[15:05:16] <SWPadnos_> ok - then I'm stumped as well ;)
[15:05:22] <les_w> I calculate 0.000416 volts/rpm
[15:05:36] <alex_joni> SWPadnos_: probably some awkward naming afterall
[15:05:49] <SWPadnos_> no - that would never happen in EMC!!
[15:05:59] <alex_joni> not once :)
[15:06:03] <les_w> haha
[15:08:59] <les_w> is the mazak set up to control the spindle speed from gcode?
[15:09:14] <alex_joni> yup, but that's emc2
[15:09:27] <les_w> I see
[15:09:42] <alex_joni> should work with emc1 too
[15:09:57] <les_w> yeah, just no documentation I guess
[15:10:22] <rayh> IMO EMC2 is the only way to run more than a minimill.
[15:11:37] <les_w> I just fear it sometimes will be broken
[15:11:52] <les_w> this machine is a poor testbed
[15:12:05] <alex_joni> les_w: then wait for a release
[15:12:26] <les_w> anything wrong and tons of steel and 10 kw of power going nuts...
[15:13:08] <les_w> I suspect the spindle speed output works fine....just no docs
[15:13:53] <alex_joni> probably
[15:13:53] <les_w> let me check the wiki
[15:13:54] <rayh> The only thing I question about spindle control with emc is the range.
[15:14:18] <rayh> You need to tell emc what s4000 means in terms of volts out.
[15:14:29] <les_w> yeah. I just need one volts per rpm parameter
[15:14:44] <les_w> the on off stuff I will run from the parport
[15:15:40] <rayh> If your controller only handles 0-10 or 4-20mv then you need direction as well
[15:15:53] <rayh> that can come from the bridgeporttask and brideportio
[15:18:29] <les_w> I will just hardwire direction at the vfd
[15:19:28] <les_w> i'll put the vfd freq output and fault output in the estop chain
[15:20:21] <les_w> I can run command voltage and speed output voltage through a comparator
[15:20:39] <rayh> If you have an "up to speed" output, you could put that as a hold on motion.
[15:20:41] <les_w> blah...checked the wiki...nothing
[15:20:54] <alex_joni> les_w: then add it after you figure it out
[15:21:06] <les_w> I was thinking just that
[15:21:12] <alex_joni> great ;)
[15:22:22] <les_w> I already have SPINDLE_ON_INDEX = 3 so it might be already working
[15:22:35] <les_w> who knows what voltage though
[15:23:06] <les_w> perhaps there are some rems in the code?
[15:23:19] <les_w> like bridgeportio.c?
[15:23:26] <les_w> let me check
[15:24:49] <SWPadnos_> there are 3 spindle indexes - SPINDLE_ON, SPINDLE_FWD and SPINDLE_REV (plus the inc and dec outputs)
[15:25:03] <SWPadnos_> those are just for controlling whether it's on or not
[15:25:21] <les_w> right...digital outs
[15:25:57] <rayh> There was an analog spindle signal as well. PWM if I remember.
[15:26:09] <rayh> but I think it was commented out in task.
[15:26:35] <SWPadnos_> no - it's there - in bridgeporttaskintf.cc
[15:27:28] <SWPadnos_> it sends the speed to the DAC index set by SPINDLE_ON_INDEX
[15:28:04] <SWPadnos_> it's treated as an axis there
[15:28:35] <SWPadnos_> ok - so SPINDLE_ON is an analog index, but FWD/REV/BRAKE are digitals
[15:29:23] <SWPadnos_> and yes - the MIN and MAX are for reverse and forward scaling, respectively
[15:29:37] <alex_joni> bleah.. bad naming :/
[15:29:38] <SWPadnos_> dacout = (speed > 0.0 ? (fabs(speed) * MAX_VOLTS_PER_RPM) :
[15:29:41] <SWPadnos_> speed < 0.0 ? (fabs(speed) * MIN_VOLTS_PER_RPM) : 0);
[15:29:44] <SWPadnos_> yes
[15:32:17] <les_w> ok
[15:32:24] <les_w> yeah very bad naming
[15:32:55] <SWPadnos_> FWD_ and REV_ might possibly have been better prefixes
[15:33:46] <rayh> phone brb
[15:34:12] <les_w> well looks like MAX_VOLTS_PER_RPM = .000416 is the thing to use
[15:34:40] <les_w> that should give S24000 a 10v output
[15:34:42] <SWPadnos_> 24 kRPM spindle?
[15:35:02] <les_w> yeah zero to 24k
[15:35:22] <les_w> the vfd goes to 1000 hz but the motor would fly apart
[15:35:47] <les_w> so I want the stg to saturate at 24000
[15:35:47] <SWPadnos_> 24k@what frequency?
[15:36:09] <les_w> 400 hz...it's a 2 pole 3 phase induction motor
[15:36:14] <SWPadnos_> wait a sec - you may be thinking backwards here
[15:36:27] <les_w> ?
[15:36:29] <les_w> haha
[15:36:34] <SWPadnos_> if you tell it that full scale (10V) is 24k RPM, then you'
[15:36:58] <SWPadnos_> if you tell it that full scale (10V) is 24k RPM, then you'll get the max available (since you'll be driving the output to full scale)
[15:37:18] <les_w> yes...
[15:37:49] <les_w> and I would hope an S12000 would put out 5 volts and so on
[15:37:58] <SWPadnos_> so you'll be setting the VFD such that full scale input is only (whatever makes 24k) output?
[15:38:41] <les_w> well the vfd is 0-10v or 4-20 ma
[15:38:54] <les_w> but 10v= 1000 hz
[15:38:59] <les_w> I want 400 max
[15:39:03] <SWPadnos_> consider this - if you tell emc that 1V = 1 RPM, and ask for 10 RPM, what will you get?
[15:39:15] <alex_joni> 10V :D
[15:39:17] <les_w> 10 I guess
[15:39:18] <SWPadnos_> answer: 10V output, and full speed on the VFD
[15:39:38] <SWPadnos_> *not* 10 RPM
[15:39:41] <alex_joni> SWPampy: where's your point?
[15:39:59] <SWPadnos_> there's no spindle speed limit in emc (that I noticed)
[15:40:09] <alex_joni> no.. but an analog limit on the DAC
[15:40:15] <alex_joni> won't go further than 10V+
[15:40:18] <les_w> The max voltage I want to the vfd is 4
[15:40:20] <alex_joni> at least on the STG
[15:40:31] <SWPadnos_> bing - there's the crucial pice of info
[15:40:36] <SWPadnos_> piece
[15:40:45] <les_w> I would put in a divider so max dac output of 10v gives 4 at the drive
[15:40:47] <SWPadnos_> 4V = 24kRPM, and 10V would be explosive RPM
[15:41:04] <SWPadnos_> ok - you're fixing it in outboard hardware
[15:41:08] <les_w> so it will never get a command that would destroy the motor
[15:41:09] <alex_joni> sounds good to me
[15:41:26] <SWPadnos_> ok - problem solved
[15:41:45] <les_w> 1000 hz= 60,000 rpm
[15:41:55] <SWPadnos_> =shrapnel
[15:41:56] <les_w> and bad bad things
[15:42:05] <les_w> heh
[15:42:19] <SWPadnos_> you may be able to set the VFD to a lower top speed - check the settings
[15:42:31] <les_w> but the dac can put out max ten...so dividing it down is the safe thing to do
[15:43:12] <SWPadnos_> a speed sensor in the estop chain is the safe thing to do
[15:43:34] <SWPadnos_> an op-amp could fail, and put full supply voltage on the output
[15:43:43] <SWPadnos_> (but that's just paranoid thinking ;) )
[15:43:56] <les_w> yes...a real speed sensor is much better than just a frequency measurement from the drive
[15:44:12] <les_w> so
[15:44:26] <les_w> I bought a melexis hall sensor
[15:45:43] <les_w> starting motion with the motor stopped is one of the worst crash scenarios
[15:45:55] <les_w> big log splitter.
[15:46:20] <SWPadnos_> unless you're just testing a program with no workpiece present (or tool, for that matter)
[15:46:36] <les_w> yeah
[15:46:57] <les_w> this thing has 10 kN force
[15:47:22] <SWPadnos_> more than I can ouput
[15:47:25] <SWPadnos_> output
[15:47:33] <les_w> 17 with all 3 axes moving (diagonal)
[15:48:54] <les_w> which is how I get a 1000 lb gantry to do like a fast plotter pen
[15:49:16] <SWPadnos_> you could almost mill a Bridgeport from a block of steel ;)
[15:49:25] <les_w> heh
[15:49:43] <les_w> well I designed it to be partly self replicating
[15:50:15] <les_w> But I am not going to cut much metal with the new spindle
[15:50:41] <les_w> the docs say "non-ferrous metals with caution"
[15:51:06] <les_w> the class 7 angulars in the thing cost $800
[15:51:42] <SWPadnos_> gah - that's a bit much for me (I just want to mill aluminum and copper)
[15:51:57] <SWPadnos_> though I may need a high speed spindle for fine work
[15:52:10] <les_w> I'll use a lower speed spindle for metal
[15:52:25] <les_w> or just use the VMC I am getting...
[15:53:28] <les_w> 4 kw is not enough power to really do high speed metal machining anyway
[15:54:25] <les_w> most pump 20 kW into a 12mm end mill!
[15:55:43] <les_w> which amazes me
[15:56:43] <les_w> anyway, thanks for the help. Guess I will just have to scope the stg dacs and see what happens
[15:57:13] <les_w> I'll make a test program with lots of S words
[15:58:11] <alex_joni> hSSSSS
[15:58:17] <alex_joni> sounds like python to me ;)
[15:58:20] <les_w> haha
[16:08:06] <Jymmm> yo
[16:08:40] <alex_joni> yo yo yo
[16:09:19] <Jacky^> hello mr Jymmm :)
[16:09:55] <Jacky^> rain .. its raining again here
[16:10:04] <Jymmm> * Jymmm stop peeing
[16:10:07] <Jacky^> like in the supertramp song !
[16:12:57] <SWPadnos_> we've got Eurythmics weather here - "Here Comes The Rain Again"
[16:13:08] <Jacky^> haha
[16:13:19] <Jymmm> we just have clouds right now
[16:13:25] <SWPadnos_> we may only be at the A-Ha stage though "Gray Skies"
[16:14:13] <Jacky^> :)
[16:14:27] <skunkworks> rain drops keep falling on my head... hmmm hmmm hmmm hmmmhmm hmmm - actually it is sunny here. don't know what the heck I'm talking about
[16:14:35] <les_w> no clouds or wind here...49 degrees f
[16:14:46] <Jymmm> ouch
[16:14:53] <SWPadnos_> yeah -I think we broke records yeterday - 67 deg or so
[16:15:09] <les_w> might get to 55 or 60 here
[16:15:12] <Jymmm> Since gas prices skyrocketed this winter, we are reluctant to light the heater.
[16:15:19] <SWPadnos_> ok - so where should the mp3 archive go? doesn't the LSB tell me that???
[16:15:35] <Jymmm> Funny thing though, lighting some candles really took off the chill.
[16:15:53] <SWPadnos_> they're especially useful for that when camping
[16:15:59] <les_w> a candle puts out a couple hundred watts
[16:16:04] <SWPadnos_> but you have to be careful to not burn down the tent
[16:16:37] <skunkworks> we keep the house at 61 degrees in the winter. although we keep it at 61 in the summer aslo so it probably evens out. ;)
[16:16:39] <Jymmm> Lit 15 and had to put out 7 as it got too warm
[16:17:09] <les_w> I am trying to use wood rather than gas
[16:17:21] <SWPadnos_> you're probably better off just turning on the heat
[16:17:26] <SWPadnos_> (rather than candles)
[16:17:48] <Jymmm> I have shitloads of candles
[16:17:57] <Jymmm> walmart has em REAL cheap too
[16:18:05] <SWPadnos_> oh - itf you're trying to get rid of them it's another story
[16:19:07] <Jymmm> It's a 2br apartment =)
[16:19:38] <les_w> I am actually low on dry wood...will have to order some. Have lots of wet wood.
[16:19:46] <Jymmm> Hell just the pilot light takes $7/mo
[16:20:36] <les_w> and more soon...a 4 ft diameter red oak is coming down to make way for the shop enlargement. It's in bad shape anyway.
[16:22:35] <les_w> wood heat works, but it's a pain. several hundred pounds a day....all loaded by hand one piece at a time
[16:22:43] <Jymmm> can't do anything with the lumber aspects les?
[16:22:55] <Jymmm> 4' diam and all
[16:23:23] <les_w> well, we will saw a nice tree...but trees in yards are usually no good
[16:23:55] <les_w> full of nails and too spread out
[16:24:01] <Jymmm> ah
[16:24:19] <les_w> even the best trees have only 20-40% usable lumber
[16:25:05] <Jymmm> les_w Oh, you said I shouldn't use a jacobs chuck, but what about if I need to drill 3/32" holes - assuming a drillbit would be better than a mill for the blind holes.
[16:25:16] <les_w> basically anything that is not vertical will be reaction wood...useless for lumber
[16:26:14] <les_w> twist drills in routers= bad
[16:26:35] <les_w> use a pcb drill bit
[16:26:56] <les_w> if you can find that size
[16:27:01] <Jymmm> * Jymmm looks....
[16:27:37] <cradek> http://www.thinktink.com/cgi-bin/carbide.cgi?page=fractional&step=20&start=10
[16:27:49] <les_w> a spiral upcut will work too if the holes are shallow
[16:28:19] <Jymmm> cradek thanks!
[16:28:25] <cradek> np
[16:28:56] <Jymmm> Now, what RPM would I need for the PCB bits?
[16:29:03] <Jymmm> into hard/soft wood
[16:29:23] <les_w> try slow
[16:29:27] <les_w> 10k or so
[16:29:36] <Jymmm> is 10K slow enough?
[16:29:58] <les_w> I drill the .25 holes in the calls at 10k
[16:30:44] <les_w> higher speeds vibrate, and have too low chip load
[16:31:07] <Jymmm> No, I just wanted to make sure that I can go SLOW enough,
[16:31:21] <les_w> I drill at 40 ipm vertical speed
[16:31:36] <Jymmm> yeah? cool
[16:32:03] <Jymmm> I guess it was a very good thing I got that 1/4 to 1/8 collet adapter =)
[16:32:04] <les_w> for deep holes use a peck cycle...g83?
[16:32:08] <les_w> forget
[16:32:24] <Jymmm> max depth 1/2"
[16:32:39] <les_w> yeah use several pecks
[16:33:00] <les_w> and as low a speed as you have
[16:33:22] <Jymmm> pecks; to clear out chips?
[16:33:26] <les_w> yeah
[16:33:29] <Jymmm> k
[16:34:49] <Jymmm> cradek have you ordered from these guys before?
[16:34:59] <cradek> yes, several times
[16:35:10] <cradek> their pcb mechanical etching tools are great
[16:35:29] <cradek> ten-packs of one size are cheaper
[16:35:51] <cradek> but they don't complain if you want one of that size, three of this size, etc
[16:36:20] <Jymmm> oh so you can mix and match to get 'ten pack' pricing?
[16:36:27] <cradek> no, I don't think so
[16:36:38] <Jymmm> no biggy.
[16:36:54] <Jymmm> will have to figure out the $50 minimum though
[16:37:26] <cradek> oh, I don't remember that
[16:37:30] <cradek> I was probably always over it
[16:38:03] <Jymmm> ah, ok
[16:38:05] <les_w> I was just thinking of how many holes emc has drilled here
[16:38:23] <les_w> I figure 80,000 or so last year
[16:39:14] <Jymmm> carbide and tungstun carbide are totally two different things, correct?
[16:39:29] <Jymmm> http://www.store.yahoo.com/drillcity/3320938.html
[16:39:54] <les_w> carbide in tools is tungsten carbide particles in a cobalt matrix
[16:40:16] <Jymmm> ah, ok.
[16:43:58] <Jymmm> depth rings CAN be removed, correct?
[16:44:04] <cradek> yes
[16:44:14] <cradek> they press off
[16:44:38] <SWPadnos_> interesting that they're free - seems quite useful
[16:44:42] <Jymmm> ok, I'd rather have them to remind me of the size... the mind is a terribele thing to waste
[16:44:58] <cradek> yeah they're good for that I suppose
[16:45:00] <Jymmm> SWPadnos my thoughts exactly
[16:45:23] <cradek> I don't use them, and I have many sizes
[16:45:25] <cradek> maybe I'll start
[16:45:29] <Jymmm> I'd hate to start a job just to realize I put in the wrong size bit
[16:45:47] <cradek> I always measure drills with the dial caliper
[16:45:52] <Jymmm> becasue I was in too much of a hurry, distracted, etc.
[16:45:54] <SWPadnos_> not precise enough for shape milling (not that yu'd do that with them anyway), but good for going through a board
[16:46:13] <SWPadnos_> they're also color-coded, I think
[16:46:19] <SWPadnos_> even easier to recognize
[16:46:20] <Jymmm> Interesting.... http://www.precisebits.com/products/scoreengrave.asp?txtsearchParamTxt=&txtsearchParamCat=20&txtsearchParamType=ALL&iLevel=1&txtsearchParamMan=ALL&txtsearchParamVen=ALL&txtFromSearch=fromSearch&btnSearch.x=10&btnSearch.y=11
[16:47:05] <cradek> Jymmm: I use those 60° tools
[16:47:17] <Jymmm> cradek what for?
[16:47:19] <SWPadnos_> hm - Zirconium Nitride coating... - wonder how good that is
[16:47:22] <cradek> pcb milling
[16:47:39] <Jymmm> cradek ok, cool. Haw you tried them on al, brass by chance?
[16:47:46] <cradek> http://timeguy.com/cradek/cnc/pcb
[16:48:15] <cradek> Jymmm: you won't be able to cut very deep in metal
[16:48:26] <cradek> they're not made for that
[16:48:32] <Jymmm> cradek: engrave
[16:48:39] <cradek> I have done that, yes
[16:48:43] <les_w> they are too brittle
[16:48:54] <cradek> no more than .01" deep
[16:49:00] <cradek> yes you can easily break the point off
[16:49:14] <Jymmm> oh, just read.... Not suitable for metal engraving
[16:49:19] <cradek> they are made to be hard for fiberglass PCBs
[16:49:19] <SWPadnos_> they don't even mention the cutter diameter - just the shank
[16:49:32] <cradek> SWPadnos_: because you only use the very tip
[16:49:40] <SWPadnos_> right
[16:49:49] <SWPadnos_> you're only *meant* to usethe tip ;)
[16:49:54] <cradek> yeah
[16:50:13] <cradek> jepler used the whole tool once on my mill
[16:50:18] <cradek> then we put a new tool in
[16:50:21] <SWPadnos_> oops
[16:50:25] <Jymmm> lol
[16:50:32] <skunkworks> Jepler - are you there? I installed the latest axis on the machine here at work - defenatly brightens up the cone. (and for some reason it is a lot brighter than at home)
[16:50:38] <SWPadnos_> keep him with the etch-a-cnc
[16:50:45] <Jymmm> ROTF
[16:50:46] <cradek> skunkworks: that's good to hear
[16:50:55] <Jymmm> brb coffee....
[16:59:11] <Jymmm> back
[16:59:13] <skunkworks> give me another week and I will be debugging code for you guys.
[16:59:18] <skunkworks> no - not really
[17:00:23] <alex_joni> skunkworks: can we settle for 2 weeks?
[17:00:58] <skunkworks> Give me 3 - I have to learn C ;)
[17:01:11] <Jymmm> skunkworks you got 30 days
[17:01:27] <Jymmm> skunkworks 29.925 days
[17:01:34] <Jymmm> skunkworks 29.8.75 days
[17:01:48] <Jymmm> (well, fucked up that joke)
[17:02:21] <SWPadnos> yeah - everyone knows there are 86400 seconds in a day
[17:02:31] <alex_joni> Jymmm: you always do :(
[17:04:09] <skunkworks> bio - I am a network admin for a small company. I mostly do vb for appication (msaccess). So I am not a total idiot - just a bit. I actually wrote and machine control program in basic that ran steppers - took a g-code file and worked pretty good (this was 10 years ago) And yes basic
[17:04:11] <rayh> Just got a note from a guy who tested the Japanese language emc file and thinks it's great.
[17:04:23] <Jymmm> alex_joni: go crawl back under the rock! =)
[17:04:41] <alex_joni> Jymmm: sorry, not enough energy to complete that
[17:04:49] <Jymmm> alex_joni: Ok, carry on =)
[17:04:50] <alex_joni> skunkworks: ewww ;) vb? :P
[17:05:00] <skunkworks> scary huh ;)
[17:07:11] <alex_joni> yeah
[17:10:42] <skunkworks> ok - I do know a little C++ - I wrote a tape emulator software for our big mill. Paper tape sucks.
[17:11:55] <alex_joni> well then.. you should get your hands dirty :)
[17:17:08] <wb9mjn> Hi All...I used the PCB bits to make the bits and pieces for the CNC conversion of the Proxon MF70 here..as the MF70 came with a 1/8 inch collet...
[17:17:39] <wb9mjn> If the bit is big enough not to break, say .125 or bigger, it will eat alluminum for lunch....and brass for dinner...
[17:19:17] <wb9mjn> I also use them at work on BeCU contacts with good success....
[17:19:27] <wb9mjn> That is hardened BeCU ....
[17:20:17] <wb9mjn> But you need to be dead on when you drill...any side load at all and they are toast...Always buy in at least packs of 5 for the small diameters...
[17:23:32] <les_w> BeCu can be pretty hard...rc40+ IIRC
[17:23:41] <wb9mjn> Yep...
[17:24:21] <les_w> I was trying to find some Be foil for some speaker drivers
[17:24:25] <wb9mjn> I usually do not drill it hardened,,,just resize holes....The stuff moves when they harden it...
[17:24:31] <les_w> ridiculous inflated prices
[17:25:03] <wb9mjn> Be careful with Be...BeCU is only 5 % Be typically, and the Be is in alloy....
[17:25:53] <les_w> I used to make phono cartrige cantelevers out of pure Be foil...fairly familiar with it
[17:26:09] <wb9mjn> 100 % Be represents a serious health risk, and should only be handled and machined in a sealed enviroment...
[17:26:21] <les_w> It has the highest speed of sound of any element
[17:26:36] <les_w> Yes, we took precautions
[17:26:36] <skunkworks> cradek - nice clocks. - I had asked jepler about the feed rate for jogging. The standard emc interface lets you select the feed rate you can run. axis only seems to let you run the feed whatever the feed override will let you. (am I making sense?)
[17:27:37] <wb9mjn> Be has allot of critical weapons applications....this is probably why you had difficulty finding it...
[17:28:31] <les_w> I think the silly prices are due to terror issues...Be is required to make a nuclear weapon
[17:29:01] <wb9mjn> Be is required to make thermo-nuclear weapons....not nuclear...
[17:29:14] <les_w> the tamper ...
[17:30:21] <rayh> skunkworks: The manual feedrate slider in tkemc is only used within the GUI
[17:30:26] <wb9mjn> That is the trick that Doctor Strangelove (Edmund Teller) came up with...
[17:30:42] <wb9mjn> anyway....
[17:30:51] <les_w> yeah
[17:30:58] <les_w> teller -ulam?
[17:31:06] <rayh> to configure the nml command for manual jog.
[17:31:16] <wb9mjn> Got my tool table EDM'd and installed....
[17:31:38] <wb9mjn> The EDM'g cost as much as the table ! About $125 for both...
[17:31:38] <alex_joni> skunkworks: the feedrate override should work on jogs too (in AXIS)
[17:31:49] <rayh> I don't know Axis
[17:32:08] <rayh> Yes alex_joni it does but what feedrate is commanded for manual jogs?
[17:32:12] <les_w> wb9mjn: picture?
[17:32:22] <alex_joni> the default speed I reckon
[17:32:33] <alex_joni> with feed override applied to that
[17:32:45] <wb9mjn> Reworking the limits, working up a gaurd for the X axis encoder, and need to figure out how to gaurd the X ways and screw...
[17:33:08] <rayh> If rapid were the commanded then feedrate override would work
[17:33:16] <les_w> I use a window shade type of thing
[17:33:32] <wb9mjn> Not yet...I can try and send it again...This lousy SBC service does not give one a website...
[17:34:01] <wb9mjn> Cloudy here still...will have to edit the picture some to make it visible, after I take it...
[17:36:13] <skunkworks> alex - yes it does but it only gives you 120% increase in the feed rate (or what ever you have the max % set to in the ini file) but the default interface has a feed rate slider for the jogging feed.
[17:36:28] <skunkworks> rate
[17:36:45] <alex_joni> so.. you could use MDI to set a speed
[17:36:49] <alex_joni> F100000
[17:36:51] <alex_joni> ;)
[17:37:27] <skunkworks> ? I don't think the last modal feed is what is used in jogging - that would be scary ;)
[17:37:35] <skunkworks> Or is it?
[17:37:46] <alex_joni> not sure.. try it :)
[17:40:46] <rayh> The easy solution for Axis is to simply use the rapid velocity for feeds
[17:41:03] <rayh> then the feedrate override will handle the chore of slowing things down.
[17:41:05] <skunkworks> It isn't - wipes sweat off brouw. ;)
[17:41:55] <skunkworks> That is a good idea ray. I really like the axis interface.
[17:42:37] <rayh> I'm not certain that I'd want rapid as the default manual jog but anything less and you can't jog at rapid.
[17:43:07] <alex_joni> surely it is done somehow in AXIS right now. maybe jepler or cradek can point us out how they thought about this
[17:43:18] <rayh> Right.
[17:43:47] <rayh> If we started it in a terminal, could we watch the manual jog NML?
[17:44:00] <alex_joni> sure
[17:44:16] <rayh> not certain I've got a running copy now.
[17:45:03] <skunkworks> what do you want to watch - I have a test system running here. i was just jogging
[17:46:36] <rayh> If you start from a terminal, do you see a command in it when you press manual jog?
[17:48:43] <rayh> if not, edit your ini to read DEBUG = 0x7FFFFFFF
[17:48:57] <rayh> restart and issue a single incremental jog
[17:49:04] <skunkworks> issuing emc_axis_jog
[17:49:41] <skunkworks> EMC_AXIS_JOG
[17:49:57] <rayh> No numbers with that?
[18:00:14] <rayh> Looks like Axis uses default jog speed as it's manual jog speed.
[18:00:53] <rayh> In default setups that is 1 inch per minute.
[18:01:53] <skunkworks> sorry - got side tracked
[18:02:05] <skunkworks> did you get it - it did have numbers behind
[18:02:18] <alex_joni> we didn't get the numbers
[18:02:48] <alex_joni> ok.. I need to go away for a while..
[18:02:53] <alex_joni> feeling worse :(
[18:02:54] <rayh> I tested here and it does use default jog speed for manual jogs.
[18:03:16] <rayh> the good news is that default jog speed isn't used anywhere else, I don't think.
[18:03:47] <rayh> so you can set it in the ini file for any max manual jog speed you like
[18:04:20] <wb9mjn> Hi Les, have some pictures now...
[18:06:39] <skunkworks> that should work for me. Just can't be in control of everything all the time. ;)
[18:07:33] <les_w> woops was away
[18:07:39] <les_w> url?
[18:08:24] <cradek> ray is right, you set the jog vel to your max vel in the ini
[18:08:28] <cradek> then use feed override to slow down
[18:08:34] <cradek> a second slider is unnecessary
[18:08:42] <cradek> skunkworks: ^^
[18:08:56] <rayh> If you set it to MAX_VELOCITY/MAX_FEED_OVERRIDE you'd get a somewhat safer manual jog
[18:09:04] <les_w> yeah g0 and jog speed are the same
[18:09:13] <rayh> and still be able to rapid if you really wanted.
[18:09:24] <les_w> only home speed is different
[18:10:17] <wb9mjn> Ok...it wont let me send due to spam problems...is there somewhere I can upload ?
[18:10:32] <SWPadnos_> what client are you using?
[18:11:02] <wb9mjn> The freenode wont let me send...My client tried ...
[18:11:31] <SWPadnos_> ah - pictures - nevermind
[18:11:50] <les_w> registered?
[18:12:37] <rayh> web9
[18:12:45] <rayh> you need to register your name
[18:13:00] <les_w> did you do a /ns identify <password>?
[18:13:10] <les_w> do that and you can prob dcc
[18:14:09] <wb9mjn> Yea...I m reading the help file...
[18:14:44] <wb9mjn> Nope...not registered here...working on it...
[18:14:59] <les_w> just pick a password and do the identify
[18:15:12] <les_w> right from the client window you are using
[18:21:40] <wb9mjn> Trying again...
[18:22:25] <wb9mjn> No connections getting made...
[18:22:39] <wb9mjn> But, no reject by the IRC server either...
[18:22:52] <wb9mjn> Might need a firewall hole...
[18:22:57] <cradek> wb9mjn: are you behind nat?
[18:23:05] <wb9mjn> nat ?
[18:23:26] <cradek> yeah
[18:23:29] <wb9mjn> IRC says the send to Les_w failed...
[18:23:32] <cradek> does your machine have a real IP address
[18:23:37] <wb9mjn> What is a "nat" ?
[18:23:42] <cradek> or is it a fake one whose traffic is translated somewhere else upstream
[18:24:05] <cradek> what's your machine's ip address?
[18:24:55] <wb9mjn> I m on a DSL link....It has a DHIP, or whatever the protocoll is that sets the addresses...the local network here has IP addressess that are not real..
[18:25:18] <wb9mjn> They get translated by SBC when the packets go into their network...
[18:25:27] <cradek> then you probably can't use dcc
[18:25:59] <wb9mjn> Ok....well that was my main gripe with SBC that they would not do a fixed IP address....
[18:26:10] <wb9mjn> Even when the sales guy said they would....
[18:26:25] <les_w> I'm not sure whether mine is fixed or not...but do dcc all the time
[18:26:34] <cradek> sometimes we have to make compromises to get cheap broadband at home
[18:26:45] <cradek> les_w: what's your computer's ip address?
[18:27:16] <wb9mjn> I have to think there is a way to do it....
[18:27:21] <les_w> I am not running a server
[18:27:25] <wb9mjn> though...
[18:27:42] <cradek> I think the translator (nat box) needs to understand irc dcc and compensate
[18:27:58] <cradek> les_w: I know - I wonder what your machine itself thinks its IP address is
[18:28:01] <les_w> let me see what address the isp is using for me
[18:28:08] <SWPadnos_> I didn't do anything special on mine, and I can dcc send/receive OK
[18:28:51] <wb9mjn> Les try to send me a .jpg and I ll see if the packets get through this way...
[18:29:03] <cradek> http://www.dslreports.com/faq/2186
[18:29:31] <cradek> of course we're not using mirc, but the idea would be the same
[18:29:58] <les_w> k
[18:30:00] <SWPadnos_> interesting - lemme check my router settings
[18:30:08] <SWPadnos_> I know I can send, and I think I've received as well
[18:30:25] <cradek> SWPadnos_: sending is no problem
[18:30:29] <cradek> receiving is the problem
[18:30:40] <cradek> ... I think
[18:30:43] <SWPadnos_> yep - that's why I want to check myself
[18:30:51] <cradek> (I have a real IP address)
[18:31:02] <SWPadnos_> well - only the Quake III ports are forwarded ;)
[18:31:09] <SWPadnos_> try sending me something
[18:31:34] <SWPadnos_> well - that worked
[18:31:40] <cradek> now send it back
[18:32:01] <cradek> the other way this can work is if your nat box knows about the dcc protocol and "fixes it up"
[18:32:18] <SWPadnos_> yeah - could be
[18:32:29] <SWPadnos_> it's a pretty recent Netgear
[18:33:10] <SWPadnos_> though I do have a static IP for the router (still NAT behind it though)
[18:33:40] <cradek> that doesn't matter
[18:33:46] <cradek> it's the nat that will mess you up
[18:33:50] <SWPadnos_> yep
[18:34:04] <cradek> I got the file but my client never said the transfer finished
[18:34:06] <cradek> did yours?
[18:34:07] <les_w> ok I seem to be
[18:34:11] <les_w> I think
[18:34:14] <SWPadnos_> I think so
[18:34:23] <SWPadnos_> I closed the tab, so I can't check
[18:34:25] <cradek> les_w: but does your computer itself have that address?
[18:34:52] <SWPadnos_> those numbers seem to be backwards
[18:34:59] <les_w> I don't think so...I just got that from the router config
[18:35:16] <cradek> les_w: sorry I wasn't clear, I was asking what your computer itself uses
[18:35:22] <SWPadnos_> you show up as leswatts@h227.116.40.69.ip.alltel.net
[18:35:30] <cradek> les_w: because if it's 192.168.x.x or 10.x.x.x you know it's using nat
[18:35:50] <cradek> % host h227.116.40.69.ip.alltel.net
[18:35:50] <cradek> h227.116.40.69.ip.alltel.net has address
[18:35:57] <les_w> weird about the backwards thing
[18:36:12] <les_w> oh
[18:36:14] <SWPadnos_> ok - I didn't see the trailing dot (or leading dot - whichever it's supposed to be)
[18:36:21] <cradek> les_w: PTR records in the DNS are written backward like that so it's normal
[18:36:24] <les_w> I know little about such things
[18:36:37] <les_w> learning though!
[18:37:24] <SWPadnos_> hm - I only knew (little) about '.xxx.in-arpa' being backwards (or something similar)
[18:37:37] <cradek> right, that's a PTR record
[18:37:43] <cradek> in-addr.arpa.
[18:37:57] <SWPadnos_> right - with the trailing period being significant
[18:37:58] <les_w> oh, of course. (huh?)
[18:37:59] <les_w> haha
[18:38:36] <cradek> les_w: a PTR record in the DNS maps from IP address to hostname
[18:38:54] <cradek> les_w: the A record goes the other way (the more important direction)
[18:39:14] <les_w> I see
[18:39:43] <wb9mjn> Tried something here...will see if it works...
[18:39:56] <jepler> cradek: so did you convince skunkworks that one slider is enough?
[18:40:05] <cradek> I'm not sure
[18:40:10] <wb9mjn> Does not look like it...
[18:40:14] <cradek> maybe I was too late and missed the conversation
[18:40:20] <les_w> got a window
[18:40:35] <cradek> wb9mjn: so you can receive, not send ... strange
[18:40:36] <wb9mjn> Got something from cradek...
[18:40:56] <les_w> I accepted but snd failed
[18:41:01] <wb9mjn> Yea,..I just opened up the "IRC" firewall ports...
[18:41:12] <wb9mjn> What did you send me cradek ?
[18:41:15] <cradek> a png
[18:41:52] <wb9mjn> Ok...axis.png...got it...
[18:42:22] <wb9mjn> Let me try to send the picture I was trying to send to Les to you...
[18:42:24] <wb9mjn> here goes...
[18:42:56] <wb9mjn> No Joy....its stuck the same way...
[18:43:05] <wb9mjn> Maybe I need to restart the IRC client ?
[18:43:11] <cradek> yeah I get the request but it doesn't transfer when I accept
[18:43:20] <les_w> same here
[18:43:34] <wb9mjn> Hmm...
[18:43:38] <cradek> wb9mjn: did you tweak your irc client somehow or just change the firewall?
[18:43:38] <Jymmm> NAT doesn't like DCC btw.
[18:43:44] <wb9mjn> I m going to try a restart...
[18:45:12] <wb9mjn> I m back...
[18:45:21] <cradek> you did not miss anything
[18:46:02] <wb9mjn> Still stuck...tried your axis file, encase mine was too large...
[18:46:09] <cradek> 13:38:45 [freenode] DCC can't connect to port 4259
[18:46:35] <wb9mjn> Ok...that s my local IP address there, not the real IP address...
[18:46:35] <cradek> your machine reports
[18:46:39] <cradek> that's not a real address
[18:46:40] <cradek> right
[18:46:42] <wb9mjn> So, the translation is not happening...
[18:46:53] <cradek> you have to convince something to tweak that reported address
[18:47:06] <cradek> maybe your irc client can do it, or maybe your router can do it if it understands the dcc protocol
[18:48:14] <les_w> dcc is good to have...
[18:48:21] <cradek> yes, very nice
[18:48:31] <les_w> btw anyone have this:
[18:48:42] <les_w> http://www.harborfreight.com/cpi/ctaf/Displayitem.taf?itemnumber=3663
[18:48:54] <Jymmm> NAT doesn't like DCC, and most dedicated ethernet routers dont do DCC passthru
[18:48:55] <les_w> wonder if it's junk
[18:49:37] <Jymmm> les_w you go thru enough blades to make your own?
[18:49:54] <les_w> I use a lot of blades, and tire of long lead times and sloppy welds
[18:50:01] <cradek> Jymmm: we know, we've been talking about it for some time now
[18:50:16] <Jymmm> ah, makes sense on the lead times
[18:50:37] <Jymmm> cradek well, I keep seeing the same conversation repeating itself...
[18:50:38] <les_w> welding does not save money
[18:50:49] <les_w> but would be convinient
[18:50:55] <les_w> have 2 band saws
[18:51:07] <Jymmm> cradek unless one bought a linksys and tossed on their own kernel
[18:51:16] <wb9mjn> I might need an updaded client...This one is from like 1999...
[18:51:31] <cradek> yeah, if you do your nat with linux or freebsd, they can fixup the traffic transparently
[18:51:37] <cradek> broadband routers aren't so clever.
[18:51:39] <Jymmm> wb9mjn What router are you using?
[18:51:46] <Jacky^> wb9mjn: I like telnet :)
[18:51:50] <Jacky^> as irc client
[18:52:26] <Jacky^> wb9mjn: kidding eh ?
[18:52:39] <Jacky^> ]:)
[18:53:01] <wb9mjn> When I started using this I was using a 44.X.X.X address...which was live on the amprnet/internet...
[18:53:15] <Jymmm> wb9mjn What router are you using?
[18:53:27] <wb9mjn> So none of these goofy how do we charge people for the internet problems...
[18:53:37] <wb9mjn> 2-wire...
[18:53:58] <cradek> apparently with irssi (the client I use) you can just specify /set dcc_port 5000 and /set dcc_own_ip a.b.c.d and then forward port 5000 in the firewall
[18:54:16] <cradek> in the nat firewall
[18:54:27] <Jymmm> wb9mjn: I meant brand/model#
[18:54:53] <cradek> if your client can't do that, try irssi - it's great
[18:55:10] <wb9mjn> The brand is 2-wire....
[18:55:17] <Jacky^> this is the mine: http://www.coyotelinux.com/products.php?Product=coyote
[18:55:50] <Jacky^> an old P133 with floppy, no HD, all in memory, web interface ;)
[18:56:18] <wb9mjn> From the "broadband reports" website, the software needs to be able to pull the real IP address from the server....
[18:56:28] <wb9mjn> I cannot see that this old software does that...
[18:56:32] <jepler> WTF? "NOTE: This product is licensed for personal and educational use only. If you would like to use our products for commercial or government use, please see the "Wolverine Firewall and VPN Server"."
[18:56:36] <Jacky^> wb9mjn: thats a cheap and good linux box
[18:56:57] <jepler> How can they license it that way, given that it's the Linux kernel?
[18:57:13] <Jacky^> I use coyote
[18:57:38] <Jacky^> never look at the source code
[18:57:53] <Jacky^> it work fine, its enough for me
[18:58:08] <Jacky^> very simpe to setup too, you can do it also from win
[18:58:43] <Jacky^> I think features are comparable to cisco commercial router
[18:59:36] <wb9mjn> I actually had a 20 MHz AT runing NOS on the Ham Radio network, prior to DSL..similar situation...
[18:59:41] <jepler> That may all be true, but it looks like they're treading on thin ice license-wise
[18:59:45] <Jymmm> wb9mjn: Yeah, no mention of irc/dcc in their KB.
[19:00:00] <Jacky^> jepler: probably
[19:00:29] <Jymmm> jepler: If the make available the kernel source code, they can do that.
[19:00:31] <Jacky^> I had not much time to setup my personal Debian box for router
[19:01:14] <Jacky^> or, I tried it and it worked very nice since the first minute (4 years im using it)
[19:02:14] <Jacky^> there was sonme other Linux Router project, but seems died ..
[19:02:45] <cradek> seems to me you could do this booted from a floppy
[19:02:56] <cradek> not sure why they need 32MB of hard disk
[19:03:19] <Jacky^> because it run all in ram
[19:03:35] <cradek> and ... they say you should have at least a 486DX/25 with two PCI network cards
[19:03:44] <cradek> for all those PCI 486 machines out there?
[19:03:47] <Jacky^> im able to run multiple h323 connections..
[19:04:18] <Jacky^> cradek: yes
[19:04:50] <Jacky^> MB,ram,floppy,supply,video card, 2 eth card.
[19:04:52] <cradek> was there such a thing? I've used EISA 486s and VESA 486s and MCA 486s
[19:05:03] <cradek> don't think I've seen a PCI 486
[19:05:21] <Jacky^> oh, I used Isa cards
[19:05:23] <Jacky^> too
[19:05:31] <Jacky^> are ok
[19:05:39] <SWPadnos_> yes - there are PCI 486es
[19:05:40] <cradek> I see
[19:05:57] <Jacky^> at the moment could be simple to find some P133
[19:06:00] <SWPadnos_> though VLB was a bit more popular at the time, I think
[19:06:17] <cradek> by gosh you're right
[19:06:32] <cradek> http://www.thegreenhouse.us/th99/m/I-L/32598.php
[19:07:06] <wb9mjn> See ya...gonna try another client...
[19:07:30] <les_w> I think i'll buy that blade welder
[19:07:37] <les_w> bet the mechanicals are junk
[19:07:58] <les_w> so i'll machine some better
[19:08:02] <cradek> sometimes chinese junk hardware is a good basis for building decent machinery
[19:09:08] <les_w> The chinese/taiwanese industrial woodworking stuff I use is fine
[19:09:49] <les_w> I don't have much chinese metalworking machines
[19:10:03] <les_w> do have lots of calipers and micrometers
[19:10:15] <les_w> they are all accurate
[19:10:19] <cradek> yeah, they're often just fine
[19:10:45] <les_w> the $5 micrometers are good to .0001
[19:10:58] <les_w> the $10 dail calipers are good to .001
[19:11:32] <les_w> some don't last long, but that's ok
[19:11:40] <les_w> don't use digital ones
[19:11:54] <cradek> yeah, I don't either
[19:11:55] <les_w> left handed. Numbers upside down.
[19:12:24] <cradek> right handed. prefer a dial.
[19:12:31] <les_w> anyway, just don't like them. I use the dial for an analog computer.
[19:12:39] <cradek> right
[19:12:49] <cradek> just like how you measure intervals with a proper clock
[19:13:19] <les_w> ah...what else do I need...big book on lap....ordering supplies
[19:13:26] <les_w> always forget something
[19:14:19] <SWPadnos_> two shiny new Opteron 244 processors?
[19:14:44] <les_w> ha
[19:14:51] <SWPadnos_> damn
[19:15:18] <les_w> the wire to hook up the vfd is $125!
[19:15:46] <Jacky^> les_w: you use some laptop ? wich brand ?
[19:16:24] <SWPadnos_> have you looked into "Mustang" wire (also called Thermacord)?
[19:16:50] <les_w> no haha the big book is the MSC catalog. 5000 pages.
[19:17:09] <les_w> mustang? no. Is it high flex?
[19:18:01] <SWPadnos_> yeah -pretty flexible, heat resistant
[19:18:19] <SWPadnos_> a friend is using some for his encoder wire, without conduit or other shielding
[19:18:21] <les_w> i'll google it
[19:18:30] <SWPadnos_> sorry -limit switch wire
[19:19:19] <SWPadnos_> made by Triangle/PWC, I think
[19:20:58] <les_w> well, my main gripe with the igus wire is that it is 4 conductor min with shield
[19:21:03] <les_w> it's fat
[19:21:13] <les_w> I only need 3 for the spindle
[19:21:31] <les_w> not much room left in the flex conduit
[19:22:30] <les_w> have to run 240/60 separate for the cooling fan too
[19:22:42] <cradek> just use some romex
[19:22:48] <les_w> haha
[19:23:08] <SWPadnos_> it's a moving gantry - flexibility is important
[19:23:23] <cradek> I was ... kidding
[19:23:28] <les_w> heh
[19:23:34] <SWPadnos_> I think copper bars would be good
[19:25:14] <les_w> now let's see also need some thomson type linear bearings...running the floating spindle dust shield on them
[19:27:53] <Jacky^> talking about cables, I remember an article abou HI-FI cables, showing how a cable of E. 3k work exactly the same of 1 E. cable.. :/
[19:28:05] <SWPadnos_> can you pick me up a Tapmatic 70TC/DC while you're ordering?
[19:28:14] <les_w> oh, ok
[19:28:15] <les_w> ha
[19:28:27] <SWPadnos_> thanks - I'll buy you lunch or something
[19:28:30] <les_w> yeah sound cables...what a racket
[19:29:25] <les_w> I just use #8 zip cord
[19:33:04] <SWPadnos_> but - can't you *hear* the oxygen in the copper?
[19:35:06] <les_w> uh...yeah....it just collapses the "soundstage"
[19:35:08] <les_w> haha
[19:35:25] <Jacky^> :P
[19:35:36] <les_w> actually #8 is pushing it
[19:35:36] <les_w> but
[19:35:42] <SWPadnos_> hm - it must be one of those symmetry things
[19:35:43] <les_w> long cable run
[19:35:53] <les_w> also
[19:35:54] <SWPadnos_> live performers collapse on the soundstage if there's no oxygen present
[19:36:07] <les_w> 3000 W RMS
[19:36:36] <les_w> in the music room I always run off to
[19:37:21] <les_w> I play it at one watt
[19:37:32] <les_w> with 5000W peaks?
[19:38:29] <les_w> uh 10 log 500o?
[19:39:17] <les_w> 37 db spl
[19:39:25] <les_w> headroom
[19:42:19] <les_w> but it won't burn you or light fires like that thing I am doing the research on in the lab
[19:54:29] <Jacky^> wb wb9mjn :)
[19:54:32] <Jacky^> solved ?
[19:54:34] <wb9mjn> I m back...have an updated version of the software..but do not think this will solve the problem...
[19:54:37] <Jymmm> * Jymmm has a Tapmatic 30 TC/DC available
[19:54:49] <wb9mjn> Probably not...Can you send me something ?
[19:55:08] <wb9mjn> So, we can see if its working as before...
[19:55:21] <les_w> I can
[19:55:29] <wb9mjn> Go ahead Les...
[19:55:35] <SWPadnos> I've got a 30 TC/DC, a 50 TC/DC, and a 90X (anyone need a 90X?)
[19:56:53] <Jacky^> 21:49 DCC SEND request sent to wb9mjn: tux.jpeg
[19:57:14] <Jacky^> 21:49 DCC sent file tux.jpeg [2kB] for wb9mjn in 00:00:01 [1.84kB/s
[19:57:26] <cradek> he could alway sreceive
[19:57:30] <cradek> couldn't send
[19:57:56] <Jacky^> oh, for that I cant help :(
[19:58:01] <Jacky^> im firewalled
[19:58:11] <cradek> wb9mjn: send to me
[19:58:28] <les_w> I am trying a send but no response so far
[19:59:04] <wb9mjn> I got the filed from Jacky, and the request from Les, but it stalled after I accepted the request from Les. Jacky's file transfer seemed to go just fine...
[19:59:07] <les_w> timed out
[19:59:33] <les_w> two at once?
[19:59:49] <les_w> let me try again
[20:00:12] <wb9mjn> Nope...Jacky's had completed I beleive, before your request...
[20:00:17] <wb9mjn> go ahead and try again Les...
[20:00:21] <Jymmm> les_w you running a ethernet router?
[20:00:46] <les_w> usb
[20:00:55] <Jacky^> Jymmm: it should not so important I think
[20:01:05] <wb9mjn> Same...stuck again...
[20:01:09] <Jacky^> youre direclty talknig to external public IP
[20:01:20] <wb9mjn> Let me try to send to Jacky...
[20:01:23] <les_w> rats
[20:01:25] <les_w> k
[20:01:26] <Jacky^> les_w: may you have ports missed on your client..
[20:01:50] <Jacky^> wb9mjn: ive all ports closed from outside at the moment
[20:01:57] <Jacky^> just th 80 http :)
[20:02:16] <les_w> I was on ethernet but storm blew it out. hooked up usb
[20:02:24] <Jymmm> The way DCC works is when you send a file to someone, you are really sending them a link to a file. Now this link points to a files on your system and they have to be able to get thru your firewall to access the file. If your router doesn't understand this, which most NAT based routers don't, yothe receiver can't get the file and the send will timeout.
[20:03:52] <Jacky^> thats is the portforward
[20:03:57] <wb9mjn> The DCC frontend seems to be broken on this version...
[20:04:12] <wb9mjn> Tried Les too...
[20:04:20] <les_w> I may have a problem too then. new hardware due to the storm
[20:04:31] <Jacky^> wb9mjn: cana you recive from les_w ?
[20:04:32] <wb9mjn> The software does not try...might be a change in the method of operation...
[20:04:38] <les_w> jacky...let me send you something
[20:04:41] <Jacky^> cant*
[20:04:41] <wb9mjn> Nope...
[20:04:57] <wb9mjn> Cradek, try to send to me now...
[20:05:04] <Jacky^> les_w: try.. but I repeat my fw block all
[20:05:25] <Jacky^> waiting for request ..
[20:05:28] <les_w> oh ok
[20:05:33] <les_w> never mind!
[20:05:45] <les_w> i'll try jymmm
[20:06:11] <Jacky^> wb9mjn: i think it dcc work on you now
[20:06:22] <Jacky^> you can receive, probably send too
[20:06:58] <wb9mjn> stuck again les...
[20:08:22] <Jymmm> les_w: I see the send request, I might have it disabled
[20:08:53] <les_w> ok
[20:09:06] <wb9mjn> Figured out the new version of the program. but its stuck trying to send....like before...
[20:09:14] <les_w> I think I am working ok here
[20:09:43] <wb9mjn> Its wierd that I can receive from others but not you Les...
[20:09:52] <les_w> yeah
[20:10:11] <Jacky^> I think ports on les_w client are missed
[20:10:11] <les_w> need to send to someone else to check
[20:11:48] <Jymmm> les_w try again
[20:11:54] <les_w> k
[20:13:20] <Jymmm> les_w ok, nm.
[20:14:21] <wb9mjn> Oh well....
[20:14:41] <les_w> nm? nautical mile?
[20:14:45] <les_w> timed out
[20:14:58] <Jymmm> les_w nm == never mind
[20:15:18] <Jymmm> les_w my client is hosed right now.
[20:15:35] <wb9mjn> send to me lymmm...
[20:15:41] <les_w> k
[20:15:51] <Jymmm> wb9mjn I can't send at all... firewall
[20:15:57] <wb9mjn> Ok...
[20:16:23] <wb9mjn> Welp...maybe another day this will get figured out....need to get busy here...
[20:17:42] <les_w> yeah...still compiling my order
[20:18:16] <Jacky^> aha, irssi have a plugin dcc resume too :D
[20:18:45] <alex_joni> hello all
[20:18:53] <Jacky^> evening alex_joni
[20:19:18] <les_w> let's seenow...that tapping head....a Komo cnc router for jymmm, oh one for jacky too....
[20:20:26] <les_w> a director's costume for cradek and alex....
[20:20:43] <Jymmm> les_w: I'll trade my existing tapping head for a 150 Watt CO2 laser and power supply
[20:20:50] <alex_joni> how about something against my flu.. I'd pick that
[20:21:07] <les_w> oh it's influenza??
[20:21:07] <alex_joni> Jymmm: they make them now in 15kW ;)
[20:21:34] <Jymmm> alex_joni 150Watt would be perfect... still can run on 120VAC circuit
[20:22:02] <alex_joni> yeah.. well you can run big machines on 120VAC ;)
[20:22:21] <alex_joni> seen a big submerged welder once, I think 1200Amps, and it was on 120AC
[20:22:38] <alex_joni> something like 150Amps from the grid.. :(
[20:22:51] <alex_joni> jepler: you around?
[20:23:20] <Jymmm> les_w Ok, so you got our christmas list... is is kindling yet?
[20:23:58] <skunkworks> always wanted to make a c02 laser. Have a 1kw laser here at work.
[20:24:26] <les_w> kindling? you want kindling? here let me go out and pick up some sticks
[20:24:49] <les_w> I built a 2 kw table long ago
[20:24:56] <Jymmm> les_w: no no no, I mean is the list you just made kindilng yet
[20:25:20] <les_w> uh yeah. it's in the stove.
[20:25:27] <les_w> haha
[20:25:29] <Jymmm> ok! lol
[20:25:58] <les_w> now that director's costume for alex and cradek...
[20:26:10] <les_w> what should it look like?
[20:26:39] <les_w> bright orange wig....round red nose...floppy shoes...?
[20:28:03] <alex_joni> floppy red shoes please
[20:28:12] <alex_joni> yellow suit :)
[20:29:52] <Jacky^> hard upgrade.. later
[20:39:57] <k4ts> hello
[20:40:11] <Jacky^> hi k4ts
[20:47:07] <jepler> alex_joni: I am paying attention now and then
[20:47:45] <alex_joni> jepler: the packet has arrived ;)
[20:47:50] <alex_joni> at least I've been told so
[20:48:01] <alex_joni> so I'm really happy for that
[20:48:11] <jepler> alex_joni: how soon till you get it from your cousin or whoever it was?
[20:48:34] <alex_joni> when he'll be here (just before christmas)
[21:23:39] <alex_joni> night all
[21:47:01] <Jacky^> ugh vmware pached and update on kernel 2.6.14 :/
[21:47:04] <Jacky^> :D
[22:07:37] <wb9mjn> FYI here is some info on DCC - http://www.mirc.org/dcchelp.html
[22:37:35] <k4ts> night all
[23:07:11] <crib> re
[23:07:11] <SWPadnos_> hi rayh
[23:07:24] <rayh> Hi Steve.
[23:07:38] <rayh> You get the cold front out there yet?
[23:07:55] <SWPadnos_> kinda - it went from high 60's yesterday to around 35 now
[23:08:17] <SWPadnos_> well - 43F according to the weather applet
[23:08:38] <SWPadnos_> have you had a chance to test the PWM ppmc driver?
[23:08:42] <rayh> It was nice here while they were getting pounded with snow south by 500 miles
[23:08:49] <SWPadnos_> heh - I love that
[23:08:52] <rayh> No not yet.
[23:08:55] <SWPadnos_> ok
[23:09:08] <rayh> I'll compile soon here and hook it up.
[23:09:08] <SWPadnos_> when did you get the card(s) you have?
[23:09:17] <rayh> month ago
[23:09:36] <SWPadnos_> ok - I'd be interested to see if they're like mine, or like jmk's
[23:09:52] <rayh> jonE was talking bout a small firmware change for the handshaking problem
[23:10:06] <rayh> Your's are different from each other?
[23:10:10] <SWPadnos_> this wasn't handshaking - we get different pulse results
[23:10:11] <SWPadnos_> yes
[23:10:23] <rayh> Really.
[23:10:46] <SWPadnos_> yep - only a problem at very high pulse rates, but still different (and neither is the same as the docs)
[23:11:21] <SWPadnos_> very high meaning in the MHz+ range
[23:11:37] <SWPadnos_> (so way faster than any reasonable encoder can go)
[23:12:17] <rayh> This pwm or stepper board?
[23:12:31] <SWPadnos_> both USC boards - step outputs
[23:12:53] <SWPadnos_> neither of us has a PWM board
[23:13:28] <rayh> okay. I'll put that one in and test.
[23:13:42] <SWPadnos_> cool - thanks. do you have a scope?
[23:13:48] <rayh> I see the updates to ppmc. compiling now.
[23:13:52] <SWPadnos_> other than HALscope :)
[23:13:54] <rayh> Nope.
[23:13:58] <SWPadnos_> bummer
[23:14:13] <rayh> Yea but it blew a powersupply a while back
[23:14:49] <SWPadnos_> ah. well, if you have an extra few $k, I'll sell you mine (so I can upgrade) :)
[23:15:21] <SWPadnos_> uh-oh - double-teaming
[23:18:52] <SWPadnos_> Do you also have a step output board?
[23:18:59] <rayh_emc2> o yea but only half a brain on each
[23:19:07] <SWPadnos_> phew
[23:19:16] <rayh_emc2> step output board?
[23:19:20] <SWPadnos_> USC
[23:19:31] <SWPadnos_> or do you have PWM only?
[23:19:33] <rayh_emc2> Yes.
[23:19:37] <rayh_emc2> I have both.
[23:20:04] <rayh_emc2> Okay I've got a usc running and it shows limits on all axes.
[23:20:08] <SWPadnos_> ok - note that I'm changing the pulse width / setup time parameters to nanoseconds, so any .hal files you make will need changes
[23:20:17] <SWPadnos_> limits hit on all axes?
[23:20:26] <rayh_emc2> Yep
[23:20:42] <SWPadnos_> is this with the stock .hal files?
[23:20:48] <rayh_emc2> Helps if I turn on usc power
[23:21:07] <SWPadnos_> hmmm - the driver shouldn't load if it's not on
[23:21:27] <rayh_emc2> No it doesn't help. Let me restart it.
[23:21:33] <SWPadnos_> it needs the board ID to figure out what pins to export
[23:22:08] <rayh_emc2> It got the id with parport power.
[23:22:32] <rayh_emc2> The lights are dimmer with only parport but they are there
[23:22:45] <SWPadnos_> interesting - on mine, the load fail light comes on, and I don't think it works
[23:22:59] <rayh_emc2> Both are dim at the same time.
[23:23:04] <rayh_emc2> brb
[23:23:06] <SWPadnos_> ok
[23:32:10] <rayh_emc2> okay. I'm looking at tkemc running the ppmc stuff.
[23:32:18] <SWPadnos_> cool
[23:32:29] <rayh_emc2> How can I test what is happening compared to yours.
[23:32:49] <SWPadnos_> well - without a scope it'll be impossible. I wouldn't worry about that
[23:33:07] <SWPadnos_> I would try jogging, and/or running 3d_chips though
[23:33:16] <rayh_emc2> Wouldn't you think the differences would be % and apply at any speed.
[23:33:25] <SWPadnos_> no - I know they aren't
[23:33:54] <SWPadnos_> I noticed the problem with a ahlscope trace of command-pos and pos-fb on an axis
[23:33:59] <SWPadnos_> halscope
[23:34:00] <rayh_emc2> 3d_chips running.
[23:34:14] <SWPadnos_> great - what are you using for PIDFF?
[23:34:30] <SWPadnos_> parameters/coefficients, I mean
[23:34:56] <rayh_emc2> ah I need some switches so it reads back pulses.
[23:35:07] <rayh_emc2> 1,2,3?
[23:35:17] <SWPadnos_> 1,2,3,4 are the ones
[23:35:38] <rayh_emc2> on. do I need to restart so it reads em?
[23:35:50] <SWPadnos_> I didn't (but you never know)
[23:37:45] <rayh_emc2> still no position feedback.
[23:38:14] <SWPadnos_> hmmm
[23:38:35] <SWPadnos_> are you using the .hal files from cvs?
[23:39:38] <rayh_emc2> yes
[23:39:49] <SWPadnos_> hm - OK
[23:40:07] <rayh_emc2> we do want switches 1-4 on?
[23:40:08] <SWPadnos_> did you try exiting, cyclung USC power, and then running again?
[23:40:27] <SWPadnos_> not normally, but if you switch them on, they'll feed back step pulses just like stepgen
[23:40:48] <SWPadnos_> oh - off on my board (switch toward the terminal strip)
[23:41:00] <rayh_emc2> yep tried power
[23:41:26] <SWPadnos_> ok - I have sw1-4 toward the edge of the board, and 5-10 toward the FPGA
[23:41:42] <SWPadnos_> and the ON label is toware the SW1 label
[23:41:46] <SWPadnos_> toward
[23:42:36] <rayh_emc2> same here with switches
[23:43:01] <SWPadnos_> ok
[23:43:18] <rayh_emc2> got out of estop.
[23:43:26] <SWPadnos_> ok
[23:43:43] <rayh_emc2> we have enables on the stepgen right.
[23:43:45] <SWPadnos_> try an MDI move - just see if the feedback changes
[23:43:55] <SWPadnos_> yes - there are now (forgot about that)
[23:44:35] <rayh_emc2> No feedback at all
[23:45:04] <SWPadnos_> do you have an encoder handy?
[23:46:55] <rayh_emc2> I'm sure I could find one.
[23:47:15] <SWPadnos_> ok - I'm testing this out as well (though with Axis)
[23:47:18] <rayh_emc2> I don't see any enable signal coming from emc.
[23:48:15] <rayh_emc2> axis.0.amp-enable-out ok it's there.
[23:48:21] <SWPadnos_> hmmm - I know the enables are there in PWM, but I'm not sure for ppmc
[23:48:42] <SWPadnos_> let me make sure I'm running with the same software - give me a minute to cvs up and compile
[23:48:48] <rayh_emc2> gotta get a bit of food. back in a bit
[23:48:52] <SWPadnos_> ok
[23:50:28] <skunkworks> what is ppmc? I assume pwm - pulse width modulation
[23:50:58] <SWPadnos_> that's the name of Jon Elson's controller boards
[23:51:04] <SWPadnos_> (or at least, some of them)
[23:51:24] <skunkworks> ah